quantization_log.json

{
  "exl3": [
    {
      "process": "exl3",
      "layer": 0,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0000309140",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.515",
      "fwd_time": "1.270",
      "(v)ram": "cuda 3.5G"
    },
    {
      "process": "exl3",
      "layer": 0,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0000449163",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.797",
      "fwd_time": "0.414",
      "(v)ram": "cuda 3.71G"
    },
    {
      "process": "exl3",
      "layer": 0,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001083960",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.484",
      "fwd_time": "1.118",
      "(v)ram": "cuda 3.75G"
    },
    {
      "process": "exl3",
      "layer": 0,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001462620",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.539",
      "fwd_time": "1.535",
      "(v)ram": "cuda 6.24G"
    },
    {
      "process": "exl3",
      "layer": 0,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001404205",
      "samples": "46025",
      "damp": "0.02500",
      "time": "35.573",
      "fwd_time": "1.535",
      "(v)ram": "cuda 6.57G"
    },
    {
      "process": "exl3",
      "layer": 0,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001023026",
      "samples": "46025",
      "damp": "0.02500",
      "time": "21.074",
      "fwd_time": "3.045",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 1,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001242660",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.391",
      "fwd_time": "0.519",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 1,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001229874",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.646",
      "fwd_time": "0.367",
      "(v)ram": "cuda 8.83G"
    },
    {
      "process": "exl3",
      "layer": 1,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002080851",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.417",
      "fwd_time": "0.987",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 1,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001577277",
      "samples": "46025",
      "damp": "0.02500",
      "time": "35.272",
      "fwd_time": "1.481",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 1,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001854992",
      "samples": "46025",
      "damp": "0.02500",
      "time": "35.474",
      "fwd_time": "1.481",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 1,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001306644",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.852",
      "fwd_time": "2.863",
      "(v)ram": "cuda 8.83G"
    },
    {
      "process": "exl3",
      "layer": 2,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001306340",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.514",
      "fwd_time": "0.634",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 2,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001280089",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.678",
      "fwd_time": "0.392",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 2,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002305954",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.383",
      "fwd_time": "1.067",
      "(v)ram": "cuda 8.82G"
    },
    {
      "process": "exl3",
      "layer": 2,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002233448",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.332",
      "fwd_time": "1.514",
      "(v)ram": "cuda 8.83G"
    },
    {
      "process": "exl3",
      "layer": 2,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002544500",
      "samples": "46025",
      "damp": "0.02500",
      "time": "35.454",
      "fwd_time": "1.514",
      "(v)ram": "cuda 8.83G"
    },
    {
      "process": "exl3",
      "layer": 2,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001308655",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.556",
      "fwd_time": "2.886",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001336568",
      "samples": "46025",
      "damp": "0.02500",
      "time": "8.192",
      "fwd_time": "0.959",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001281710",
      "samples": "46025",
      "damp": "0.02500",
      "time": "13.850",
      "fwd_time": "0.959",
      "(v)ram": "cuda 9.94G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0000861670",
      "samples": "46025",
      "damp": "0.02500",
      "time": "16.600",
      "fwd_time": "0.959",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001134306",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.467",
      "fwd_time": "0.706",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002259402",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.497",
      "fwd_time": "1.143",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002669251",
      "samples": "46025",
      "damp": "0.02500",
      "time": "35.421",
      "fwd_time": "1.143",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 3,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001813367",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.349",
      "fwd_time": "2.515",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 4,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001542146",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.852",
      "fwd_time": "0.303",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 4,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001483606",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.244",
      "fwd_time": "0.336",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 4,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003873379",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.183",
      "fwd_time": "0.940",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 4,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003108313",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.741",
      "fwd_time": "1.381",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 4,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002699154",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.256",
      "fwd_time": "1.381",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 4,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002329501",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.218",
      "fwd_time": "2.732",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 5,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001508007",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.648",
      "fwd_time": "0.554",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 5,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001393725",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.212",
      "fwd_time": "0.325",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 5,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003172869",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.033",
      "fwd_time": "0.990",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 5,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002831546",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.773",
      "fwd_time": "1.326",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 5,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003238054",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.256",
      "fwd_time": "1.326",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 5,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002310051",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.291",
      "fwd_time": "2.662",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 6,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001442891",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.682",
      "fwd_time": "0.508",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 6,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001328030",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.201",
      "fwd_time": "0.328",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 6,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002876523",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.169",
      "fwd_time": "0.985",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 6,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002415747",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.799",
      "fwd_time": "1.394",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 6,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002349030",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.334",
      "fwd_time": "1.394",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 6,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001824118",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.233",
      "fwd_time": "2.757",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001386507",
      "samples": "46025",
      "damp": "0.02500",
      "time": "11.121",
      "fwd_time": "0.816",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001359356",
      "samples": "46025",
      "damp": "0.02500",
      "time": "11.278",
      "fwd_time": "0.816",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0000912586",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.375",
      "fwd_time": "0.816",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001939791",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.002",
      "fwd_time": "0.632",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003396732",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.600",
      "fwd_time": "1.038",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002888420",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.141",
      "fwd_time": "1.038",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 7,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002696260",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.243",
      "fwd_time": "2.374",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 8,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001510310",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.753",
      "fwd_time": "0.502",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 8,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001429712",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.188",
      "fwd_time": "0.332",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 8,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004271022",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.047",
      "fwd_time": "1.013",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 8,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003643227",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.668",
      "fwd_time": "1.341",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 8,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003252976",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.241",
      "fwd_time": "1.341",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 8,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002908405",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.242",
      "fwd_time": "2.628",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 9,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001559797",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.649",
      "fwd_time": "0.574",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 9,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001376020",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.167",
      "fwd_time": "0.330",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 9,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003536414",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.028",
      "fwd_time": "0.972",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 9,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003586138",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.509",
      "fwd_time": "1.326",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 9,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003178644",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.163",
      "fwd_time": "1.326",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 9,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002884370",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.265",
      "fwd_time": "2.618",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 10,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001559481",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.690",
      "fwd_time": "0.541",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 10,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001355222",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.248",
      "fwd_time": "0.352",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 10,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002927167",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.009",
      "fwd_time": "0.979",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 10,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002931710",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.495",
      "fwd_time": "1.403",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 10,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003208519",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.120",
      "fwd_time": "1.403",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 10,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002370070",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.223",
      "fwd_time": "2.686",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001399395",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.168",
      "fwd_time": "0.780",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001380504",
      "samples": "46025",
      "damp": "0.02500",
      "time": "12.025",
      "fwd_time": "0.780",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0000854362",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.457",
      "fwd_time": "0.780",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001955680",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.045",
      "fwd_time": "0.630",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002686702",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.275",
      "fwd_time": "1.059",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003034372",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.983",
      "fwd_time": "1.059",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 11,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002896157",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.277",
      "fwd_time": "2.378",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 12,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001494801",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.665",
      "fwd_time": "0.504",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 12,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001368212",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.162",
      "fwd_time": "0.338",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 12,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003907182",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.087",
      "fwd_time": "0.916",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 12,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003022966",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.648",
      "fwd_time": "1.401",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 12,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002632661",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.161",
      "fwd_time": "1.401",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 12,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003113969",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.255",
      "fwd_time": "2.685",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 13,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001520758",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.636",
      "fwd_time": "0.487",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 13,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001381202",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.230",
      "fwd_time": "0.333",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 13,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004113755",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.036",
      "fwd_time": "0.965",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 13,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002692790",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.475",
      "fwd_time": "1.397",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 13,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003085764",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.228",
      "fwd_time": "1.397",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 13,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003289077",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.228",
      "fwd_time": "2.763",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 14,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001477030",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.647",
      "fwd_time": "0.611",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 14,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001321305",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.127",
      "fwd_time": "0.325",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 14,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004008792",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.102",
      "fwd_time": "0.943",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 14,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003222023",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.367",
      "fwd_time": "1.365",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 14,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002806892",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.107",
      "fwd_time": "1.365",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 14,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003319602",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.193",
      "fwd_time": "2.737",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001635365",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.371",
      "fwd_time": "0.841",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001574359",
      "samples": "46025",
      "damp": "0.02500",
      "time": "14.209",
      "fwd_time": "0.841",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0000965914",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.393",
      "fwd_time": "0.841",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002419121",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.981",
      "fwd_time": "0.648",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002854773",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.591",
      "fwd_time": "1.077",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003213659",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.170",
      "fwd_time": "1.077",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 15,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003845903",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.049",
      "fwd_time": "2.370",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 16,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001654009",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.681",
      "fwd_time": "0.516",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 16,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001445589",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.138",
      "fwd_time": "0.326",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 16,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005057470",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.979",
      "fwd_time": "0.965",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 16,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002810189",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.301",
      "fwd_time": "1.386",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 16,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003197273",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.024",
      "fwd_time": "1.386",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 16,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003645015",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.217",
      "fwd_time": "2.647",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 17,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001624689",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.594",
      "fwd_time": "0.509",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 17,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001350336",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.177",
      "fwd_time": "0.348",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 17,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004959884",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.030",
      "fwd_time": "0.986",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 17,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002914195",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.804",
      "fwd_time": "1.364",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 17,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003383347",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.259",
      "fwd_time": "1.364",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 17,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003485196",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.283",
      "fwd_time": "2.709",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 18,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001615313",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.664",
      "fwd_time": "0.539",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 18,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001354064",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.128",
      "fwd_time": "0.325",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 18,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003866375",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.092",
      "fwd_time": "0.998",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 18,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002605821",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.933",
      "fwd_time": "1.295",
      "(v)ram": "cuda 9.99G"
    },
    {
      "process": "exl3",
      "layer": 18,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002789636",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.449",
      "fwd_time": "1.295",
      "(v)ram": "cuda 9.99G"
    },
    {
      "process": "exl3",
      "layer": 18,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002689808",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.672",
      "fwd_time": "2.615",
      "(v)ram": "cuda 9.99G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001869606",
      "samples": "46025",
      "damp": "0.02500",
      "time": "4.828",
      "fwd_time": "0.844",
      "(v)ram": "cuda 9.98G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001792437",
      "samples": "46025",
      "damp": "0.02500",
      "time": "13.130",
      "fwd_time": "0.844",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0000884415",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.422",
      "fwd_time": "0.844",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003210720",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.987",
      "fwd_time": "0.657",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003511447",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.395",
      "fwd_time": "1.036",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002796398",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.038",
      "fwd_time": "1.036",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 19,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003248990",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.276",
      "fwd_time": "2.386",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 20,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001489779",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.657",
      "fwd_time": "0.504",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 20,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001312913",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.187",
      "fwd_time": "0.338",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 20,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002977846",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.006",
      "fwd_time": "1.022",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 20,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002646870",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.478",
      "fwd_time": "1.410",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 20,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004055389",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.090",
      "fwd_time": "1.410",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 20,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002150426",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.353",
      "fwd_time": "2.631",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 21,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001719459",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.649",
      "fwd_time": "0.494",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 21,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001451151",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.193",
      "fwd_time": "0.334",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 21,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004431289",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.004",
      "fwd_time": "0.964",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 21,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004651770",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.625",
      "fwd_time": "1.338",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 21,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003120091",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.326",
      "fwd_time": "1.338",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 21,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003685448",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.060",
      "fwd_time": "2.716",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 22,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002048398",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.595",
      "fwd_time": "0.570",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 22,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001587088",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.200",
      "fwd_time": "0.347",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 22,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005549812",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.030",
      "fwd_time": "1.024",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 22,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004814410",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.484",
      "fwd_time": "1.355",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 22,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002960062",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.051",
      "fwd_time": "1.355",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 22,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004177599",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.174",
      "fwd_time": "2.681",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001928436",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.388",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0002173929",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.121",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0001174588",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.355",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003859055",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.986",
      "fwd_time": "0.635",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003295268",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.277",
      "fwd_time": "1.030",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005289865",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.032",
      "fwd_time": "1.030",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 23,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004518460",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.062",
      "fwd_time": "2.358",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 24,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002148488",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.599",
      "fwd_time": "0.460",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 24,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001902070",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.156",
      "fwd_time": "0.330",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 24,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0006128301",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.044",
      "fwd_time": "0.935",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 24,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005080805",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.560",
      "fwd_time": "1.345",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 24,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003113936",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.196",
      "fwd_time": "1.345",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 24,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004529007",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.167",
      "fwd_time": "2.714",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 25,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0001998674",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.630",
      "fwd_time": "0.570",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 25,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001677332",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.185",
      "fwd_time": "0.329",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 25,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005435301",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.037",
      "fwd_time": "0.986",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 25,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003176464",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.281",
      "fwd_time": "1.347",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 25,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004661248",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.097",
      "fwd_time": "1.347",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 25,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004602059",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.405",
      "fwd_time": "2.710",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 26,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002099135",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.635",
      "fwd_time": "0.562",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 26,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001774894",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.168",
      "fwd_time": "0.334",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 26,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005245441",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.045",
      "fwd_time": "0.929",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 26,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003808236",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.242",
      "fwd_time": "1.379",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 26,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002898486",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.067",
      "fwd_time": "1.379",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 26,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004453391",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.222",
      "fwd_time": "2.689",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001860403",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.168",
      "fwd_time": "0.842",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0001722349",
      "samples": "46025",
      "damp": "0.02500",
      "time": "13.100",
      "fwd_time": "0.842",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0001472804",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.533",
      "fwd_time": "0.842",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005299116",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.081",
      "fwd_time": "0.642",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003056609",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.638",
      "fwd_time": "1.050",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004288397",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.104",
      "fwd_time": "1.050",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 27,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005444297",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.905",
      "fwd_time": "2.383",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 28,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002335945",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.643",
      "fwd_time": "0.564",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 28,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002048673",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.185",
      "fwd_time": "0.329",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 28,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0007258357",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.034",
      "fwd_time": "0.962",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 28,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004331741",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.419",
      "fwd_time": "1.313",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 28,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003322001",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.060",
      "fwd_time": "1.313",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 28,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006044699",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.775",
      "fwd_time": "2.615",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 29,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002503210",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.688",
      "fwd_time": "0.522",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 29,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002228845",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.221",
      "fwd_time": "0.325",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 29,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008297868",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.994",
      "fwd_time": "1.100",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 29,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004543384",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.303",
      "fwd_time": "1.345",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 29,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003561271",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.043",
      "fwd_time": "1.345",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 29,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006406601",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.180",
      "fwd_time": "2.637",
      "(v)ram": "cuda 9.99G"
    },
    {
      "process": "exl3",
      "layer": 30,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002445520",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.601",
      "fwd_time": "0.620",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 30,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002153087",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.170",
      "fwd_time": "0.331",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 30,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008040649",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.019",
      "fwd_time": "0.990",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 30,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004319072",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.410",
      "fwd_time": "1.357",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 30,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003641952",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.991",
      "fwd_time": "1.357",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 30,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006349357",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.347",
      "fwd_time": "2.675",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0002497700",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.968",
      "fwd_time": "0.899",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0002127857",
      "samples": "46025",
      "damp": "0.02500",
      "time": "14.895",
      "fwd_time": "0.899",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0002150357",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.290",
      "fwd_time": "0.899",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0009070649",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.996",
      "fwd_time": "0.642",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003914292",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.237",
      "fwd_time": "1.029",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004679201",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.977",
      "fwd_time": "1.029",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 31,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006958458",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.066",
      "fwd_time": "2.380",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 32,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003037939",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.618",
      "fwd_time": "0.512",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 32,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002704197",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.183",
      "fwd_time": "0.349",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 32,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0009363733",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.100",
      "fwd_time": "0.975",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 32,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004212085",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.249",
      "fwd_time": "1.351",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 32,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005144451",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.938",
      "fwd_time": "1.351",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 32,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007600126",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.320",
      "fwd_time": "2.756",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 33,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003252792",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.613",
      "fwd_time": "0.585",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 33,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002617861",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.133",
      "fwd_time": "0.336",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 33,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0010403561",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.065",
      "fwd_time": "0.922",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 33,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005620538",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.674",
      "fwd_time": "1.451",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 33,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004516229",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.110",
      "fwd_time": "1.451",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 33,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007879546",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.728",
      "fwd_time": "2.673",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 34,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003295215",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.616",
      "fwd_time": "0.453",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 34,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002876067",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.150",
      "fwd_time": "0.333",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 34,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008338922",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.122",
      "fwd_time": "0.966",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 34,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005208213",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.548",
      "fwd_time": "1.362",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 34,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004158991",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.196",
      "fwd_time": "1.362",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 34,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007042662",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.813",
      "fwd_time": "2.688",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003362718",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.270",
      "fwd_time": "0.888",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003265047",
      "samples": "46025",
      "damp": "0.02500",
      "time": "14.284",
      "fwd_time": "0.888",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0003263641",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.546",
      "fwd_time": "0.888",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0007288093",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.030",
      "fwd_time": "0.638",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006125391",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.304",
      "fwd_time": "1.049",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003933221",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.190",
      "fwd_time": "1.049",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 35,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006420517",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.636",
      "fwd_time": "2.370",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 36,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003274105",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.770",
      "fwd_time": "0.577",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 36,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002907135",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.122",
      "fwd_time": "0.332",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 36,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0007648074",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.044",
      "fwd_time": "0.951",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 36,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004268119",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.664",
      "fwd_time": "1.353",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 36,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007536931",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.190",
      "fwd_time": "1.353",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 36,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007601988",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.967",
      "fwd_time": "2.715",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 37,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003234647",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.706",
      "fwd_time": "0.512",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 37,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002590321",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.147",
      "fwd_time": "0.328",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 37,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008620319",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.020",
      "fwd_time": "0.936",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 37,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004900083",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.418",
      "fwd_time": "1.400",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 37,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0008112190",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.064",
      "fwd_time": "1.400",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 37,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0008626336",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.042",
      "fwd_time": "2.754",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 38,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003802990",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.678",
      "fwd_time": "0.471",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 38,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003073446",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.129",
      "fwd_time": "0.355",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 38,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0010235628",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.027",
      "fwd_time": "0.967",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 38,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004016822",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.305",
      "fwd_time": "1.349",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 38,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007591551",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.104",
      "fwd_time": "1.349",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 38,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007456726",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.490",
      "fwd_time": "2.661",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003232131",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.793",
      "fwd_time": "0.929",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003390348",
      "samples": "46025",
      "damp": "0.02500",
      "time": "12.194",
      "fwd_time": "0.929",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0002820576",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.240",
      "fwd_time": "0.929",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008003408",
      "samples": "46025",
      "damp": "0.02500",
      "time": "5.967",
      "fwd_time": "0.628",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0008464451",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.630",
      "fwd_time": "1.048",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004246982",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.076",
      "fwd_time": "1.048",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 39,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007433402",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.051",
      "fwd_time": "2.374",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 40,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003490076",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.650",
      "fwd_time": "0.566",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 40,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003084388",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.133",
      "fwd_time": "0.336",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 40,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0007707802",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.025",
      "fwd_time": "0.927",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 40,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007658285",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.403",
      "fwd_time": "1.373",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 40,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004231455",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.103",
      "fwd_time": "1.373",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 40,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007200944",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.541",
      "fwd_time": "2.708",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 41,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003405369",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.634",
      "fwd_time": "0.469",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 41,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002900135",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.213",
      "fwd_time": "0.340",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 41,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008328006",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.003",
      "fwd_time": "0.977",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 41,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004961537",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.494",
      "fwd_time": "1.394",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 41,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007546357",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.216",
      "fwd_time": "1.394",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 41,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007851818",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.694",
      "fwd_time": "2.695",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 42,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003357655",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.666",
      "fwd_time": "0.498",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 42,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002920112",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.171",
      "fwd_time": "0.337",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 42,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0006470478",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.088",
      "fwd_time": "0.930",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 42,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004897234",
      "samples": "46025",
      "damp": "0.02500",
      "time": "32.460",
      "fwd_time": "1.383",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 42,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006049514",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.156",
      "fwd_time": "1.383",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 42,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006864230",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.571",
      "fwd_time": "2.648",
      "(v)ram": "cuda 10.08G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003178129",
      "samples": "46025",
      "damp": "0.02500",
      "time": "7.207",
      "fwd_time": "0.912",
      "(v)ram": "cuda 10.02G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0002634857",
      "samples": "46025",
      "damp": "0.02500",
      "time": "12.997",
      "fwd_time": "0.912",
      "(v)ram": "cuda 10.09G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0002652057",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.752",
      "fwd_time": "0.912",
      "(v)ram": "cuda 10.09G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008728239",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.106",
      "fwd_time": "0.647",
      "(v)ram": "cuda 10.G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006823797",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.492",
      "fwd_time": "1.043",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004830709",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.067",
      "fwd_time": "1.043",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 43,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0008159428",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.819",
      "fwd_time": "2.390",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 44,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003474322",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.693",
      "fwd_time": "0.598",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 44,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002955515",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.157",
      "fwd_time": "0.330",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 44,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0009270963",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.055",
      "fwd_time": "0.899",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 44,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006885539",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.635",
      "fwd_time": "1.302",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 44,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005086279",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.252",
      "fwd_time": "1.302",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 44,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0008795656",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.409",
      "fwd_time": "2.639",
      "(v)ram": "cuda 9.97G"
    },
    {
      "process": "exl3",
      "layer": 45,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003841739",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.899",
      "fwd_time": "0.473",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 45,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003138318",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.338",
      "fwd_time": "0.345",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 45,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008431417",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.215",
      "fwd_time": "0.973",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 45,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006991468",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.492",
      "fwd_time": "1.413",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 45,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005229581",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.192",
      "fwd_time": "1.413",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 45,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0008932385",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.644",
      "fwd_time": "2.780",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 46,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003554225",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.024",
      "fwd_time": "0.628",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 46,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002976912",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.396",
      "fwd_time": "0.346",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 46,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008711365",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.206",
      "fwd_time": "0.994",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 46,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005050883",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.218",
      "fwd_time": "1.430",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 46,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006381671",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.207",
      "fwd_time": "1.430",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 46,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007903029",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.831",
      "fwd_time": "2.698",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003274018",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.375",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003669015",
      "samples": "46025",
      "damp": "0.02500",
      "time": "13.757",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0003379878",
      "samples": "46025",
      "damp": "0.02500",
      "time": "16.056",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008612994",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.243",
      "fwd_time": "0.669",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006825118",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.461",
      "fwd_time": "1.081",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005044094",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.120",
      "fwd_time": "1.081",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 47,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007596755",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.696",
      "fwd_time": "2.426",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 48,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003813490",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.971",
      "fwd_time": "0.566",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 48,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003502074",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.358",
      "fwd_time": "0.343",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 48,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0006510173",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.234",
      "fwd_time": "1.031",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 48,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004991873",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.140",
      "fwd_time": "1.474",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 48,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006683612",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.135",
      "fwd_time": "1.474",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 48,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0007584409",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.672",
      "fwd_time": "2.691",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 49,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003978618",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.733",
      "fwd_time": "0.596",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 49,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003070718",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.306",
      "fwd_time": "0.344",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 49,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0008880676",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.166",
      "fwd_time": "0.927",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 49,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006291384",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.672",
      "fwd_time": "1.410",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 49,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004871004",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.143",
      "fwd_time": "1.410",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 49,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006890490",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.933",
      "fwd_time": "2.760",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 50,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003741637",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.972",
      "fwd_time": "0.523",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 50,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002721552",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.378",
      "fwd_time": "0.351",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 50,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0006299870",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.272",
      "fwd_time": "0.984",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 50,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004704391",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.500",
      "fwd_time": "1.392",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 50,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004002026",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.269",
      "fwd_time": "1.392",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 50,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005769877",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.907",
      "fwd_time": "2.696",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003784354",
      "samples": "46025",
      "damp": "0.02500",
      "time": "8.422",
      "fwd_time": "0.814",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003511167",
      "samples": "46025",
      "damp": "0.02500",
      "time": "12.476",
      "fwd_time": "0.814",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0003761418",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.920",
      "fwd_time": "0.814",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005472815",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.213",
      "fwd_time": "0.663",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004222643",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.444",
      "fwd_time": "1.089",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005779060",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.075",
      "fwd_time": "1.089",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 51,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005812909",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.086",
      "fwd_time": "2.434",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 52,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003348201",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.928",
      "fwd_time": "0.511",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 52,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003026903",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.338",
      "fwd_time": "0.354",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 52,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005606964",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.160",
      "fwd_time": "0.950",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 52,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005932474",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.963",
      "fwd_time": "1.414",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 52,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004069592",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.285",
      "fwd_time": "1.414",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 52,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006731073",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.836",
      "fwd_time": "2.710",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 53,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003404680",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.855",
      "fwd_time": "0.626",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 53,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003365407",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.374",
      "fwd_time": "0.367",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 53,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0007645086",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.229",
      "fwd_time": "1.023",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 53,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005782370",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.249",
      "fwd_time": "1.425",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 53,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003983052",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.027",
      "fwd_time": "1.425",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 53,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006790817",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.620",
      "fwd_time": "2.774",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 54,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0004027427",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.923",
      "fwd_time": "0.579",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 54,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003152286",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.312",
      "fwd_time": "0.342",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 54,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005548156",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.213",
      "fwd_time": "0.956",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 54,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003866323",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.496",
      "fwd_time": "1.439",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 54,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003268360",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.090",
      "fwd_time": "1.439",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 54,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004398462",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.777",
      "fwd_time": "2.790",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003050208",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.317",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003773301",
      "samples": "46025",
      "damp": "0.02500",
      "time": "14.671",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0003309635",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.869",
      "fwd_time": "0.925",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005151500",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.201",
      "fwd_time": "0.669",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003558416",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.327",
      "fwd_time": "1.122",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006968865",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.140",
      "fwd_time": "1.122",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 55,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004789237",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.618",
      "fwd_time": "2.446",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 56,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003393677",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.955",
      "fwd_time": "0.613",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 56,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003519310",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.366",
      "fwd_time": "0.389",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 56,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005874291",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.207",
      "fwd_time": "0.946",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 56,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006979220",
      "samples": "46025",
      "damp": "0.02500",
      "time": "30.038",
      "fwd_time": "1.410",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 56,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003509869",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.021",
      "fwd_time": "1.410",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 56,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006440084",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.712",
      "fwd_time": "2.781",
      "(v)ram": "cuda 10.03G"
    },
    {
      "process": "exl3",
      "layer": 57,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003031982",
      "samples": "46025",
      "damp": "0.02500",
      "time": "10.145",
      "fwd_time": "0.314",
      "(v)ram": "cuda 10.G"
    },
    {
      "process": "exl3",
      "layer": 57,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003348848",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.384",
      "fwd_time": "0.348",
      "(v)ram": "cuda 10.G"
    },
    {
      "process": "exl3",
      "layer": 57,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0007107153",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.234",
      "fwd_time": "0.947",
      "(v)ram": "cuda 10.G"
    },
    {
      "process": "exl3",
      "layer": 57,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003246915",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.287",
      "fwd_time": "1.360",
      "(v)ram": "cuda 9.98G"
    },
    {
      "process": "exl3",
      "layer": 57,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006968614",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.062",
      "fwd_time": "1.360",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 57,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0006694435",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.829",
      "fwd_time": "2.822",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 58,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003729765",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.918",
      "fwd_time": "0.644",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 58,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003601975",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.350",
      "fwd_time": "0.354",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 58,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0006460066",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.208",
      "fwd_time": "1.013",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 58,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003046432",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.189",
      "fwd_time": "1.453",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 58,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0005972558",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.000",
      "fwd_time": "1.453",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 58,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003008318",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.075",
      "fwd_time": "2.762",
      "(v)ram": "cuda 9.96G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0002824474",
      "samples": "46025",
      "damp": "0.02500",
      "time": "8.420",
      "fwd_time": "0.841",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003532591",
      "samples": "46025",
      "damp": "0.02500",
      "time": "14.967",
      "fwd_time": "0.841",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0003260664",
      "samples": "46025",
      "damp": "0.02500",
      "time": "15.903",
      "fwd_time": "0.841",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0003682591",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.206",
      "fwd_time": "0.661",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002810003",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.144",
      "fwd_time": "1.084",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004985353",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.058",
      "fwd_time": "1.084",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 59,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002992451",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.278",
      "fwd_time": "2.436",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 60,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002428644",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.904",
      "fwd_time": "0.574",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 60,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002445248",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.313",
      "fwd_time": "0.343",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 60,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004005747",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.140",
      "fwd_time": "0.957",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 60,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004117131",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.339",
      "fwd_time": "1.446",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 60,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002569725",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.083",
      "fwd_time": "1.446",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 60,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0004263907",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.864",
      "fwd_time": "2.745",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 61,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0003178433",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.982",
      "fwd_time": "0.626",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 61,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002764839",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.389",
      "fwd_time": "0.353",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 61,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0005426502",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.217",
      "fwd_time": "0.941",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 61,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002287043",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.515",
      "fwd_time": "1.434",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 61,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003444905",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.347",
      "fwd_time": "1.434",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 61,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0003469393",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.967",
      "fwd_time": "2.781",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 62,
      "module": "linear_attn.in_proj_qkv",
      "feat: in, out": "5120, 10240",
      "dtype: size": "bf16: 125.0MB",
      "loss": "0.0002329481",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.963",
      "fwd_time": "0.526",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 62,
      "module": "linear_attn.in_proj_z",
      "feat: in, out": "5120, 6144",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0001759246",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.326",
      "fwd_time": "0.344",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 62,
      "module": "linear_attn.out_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0004117934",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.246",
      "fwd_time": "1.038",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 62,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002198772",
      "samples": "46025",
      "damp": "0.02500",
      "time": "33.941",
      "fwd_time": "1.422",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 62,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002022106",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.220",
      "fwd_time": "1.422",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 62,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0002280657",
      "samples": "46025",
      "damp": "0.02500",
      "time": "19.970",
      "fwd_time": "2.746",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "self_attn.k_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0003487052",
      "samples": "46025",
      "damp": "0.02500",
      "time": "9.657",
      "fwd_time": "0.979",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "self_attn.v_proj",
      "feat: in, out": "5120, 1024",
      "dtype: size": "bf16: 12.5MB",
      "loss": "0.0000905891",
      "samples": "46025",
      "damp": "0.02500",
      "time": "13.704",
      "fwd_time": "0.979",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "self_attn.q_proj",
      "feat: in, out": "5120, 12288",
      "dtype: size": "bf16: 150.0MB",
      "loss": "0.0001979654",
      "samples": "46025",
      "damp": "0.02500",
      "time": "16.014",
      "fwd_time": "0.979",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "self_attn.o_proj",
      "feat: in, out": "6144, 5120",
      "dtype: size": "bf16: 75.0MB",
      "loss": "0.0002509987",
      "samples": "46025",
      "damp": "0.02500",
      "time": "6.192",
      "fwd_time": "0.659",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "mlp.gate_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001300834",
      "samples": "46025",
      "damp": "0.02500",
      "time": "31.494",
      "fwd_time": "1.113",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "mlp.up_proj",
      "feat: in, out": "5120, 17408",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0000878490",
      "samples": "46025",
      "damp": "0.02500",
      "time": "34.222",
      "fwd_time": "1.113",
      "(v)ram": "cuda 9.95G"
    },
    {
      "process": "exl3",
      "layer": 63,
      "module": "mlp.down_proj",
      "feat: in, out": "17408, 5120",
      "dtype: size": "bf16: 212.5MB",
      "loss": "0.0001124469",
      "samples": "46025",
      "damp": "0.02500",
      "time": "20.233",
      "fwd_time": "2.436",
      "(v)ram": "cuda 9.95G"
    }
  ]
}